Agent-Based Modeling as a Bridge Between Cognitive and Social Perspectives on Learning
Abrahamson, Wilensky, & Levin (2007) Agent-Based Modeling as a Bridge Between Cognitive and Social Perspectives on Learning
Dor Abrahamson, University of California - Berkeley
Uri Wilensky , Northwestern University
James Levin , University of California - San Diego
This paper is a proof-of-existence empirical paper. That said, it is also a methodological paper. The methodology is agent-based modeling, a computer-supported mode of inquiry into complex phenomena, such as weather fronts, market fluctuations, or participation patterns in a middle-school mathematics lesson. We have previously shown that agent-based simulation can express theoretical models of learning (Abrahamson & Wilensky, 2005; see also Smith & Conrey, 2007). In that paper, we claimed that a promising attribute of simulation-based research into learning is that it fosters scholarly critique and engaging collaboration. It is that claim that is herein proven to stand. The proof lies in the collaborative criticism offered by the third author, Jim Levin, to the first two authors, Dor Abrahamson and Uri Wilensky, concerning their agent-based
simulation of learning, which was first presented at the 2005 annual meeting of the Jean Piaget Society and was then made available online, along with its underlying computational procedures that were laid out for scrutiny. The 2005 paper explicitly invited fellow researchers to critique the simulation and possibly modify it so as to accommodate their own perspectives and possibly enable their own investigations. Indeed, the critique received from Levin took a unique form—he improved the computer-based model such that it better simulates the target constructs. Such co-constructive critique, we argue, is a hallmark of the promise of agent-based modeling. So we submit that this proof of existence, if anecdotal validation, may be a harbinger of a new mode of research in the learning sciences and beyond—a mode that builds on constructionism (Papert, 1991): constructionist collaboration .
Background
We have argued (Abrahamson & Wilensky, 2005) that ABM has potential to contribute to the advancement of theory in at least three major ways: (a) explicitizing—ABM computational environments demand an exacting level of clarity and specificity in expressing a theoretical model and provide the tools, structures, and standard practices to achieve this high level; (b) emergence—the computational power of ABM enables the researcher to mobilize an otherwise static list of conjectured behaviors and witness any group-level patterns that may enfold through multiple interactions between the agents who implement these conjectured behaviors; and (c) intra/inter-disciplinary collaboration—the lingua franca of ABM enables researchers, who
otherwise use different frameworks, terminology, and methodologies, to understand and critique each others’ theory and even challenge or improve the theory by modifying and/or extending the computational procedures that underlie the model. It is the latter attribute of ABM that enabled the third author to readily engage in discourse with the first two authors, following their initial presentation of the “I’m Game!” Piagetian–Vygotskiian model (Abrahamson & Wilensky, 2005).
In this paper, we begin by recapping the Abrahamson–Wilensky paper and model (hence, A–W) and then present Levin’s proposed improvement on the A–W model as a case study of the A–W call for model-based constructive critique. We end with a cautionary remark on the limitations of ABM, the importance of recognizing these limitations, and the relevance of this caution for the field’s understanding, and prospective incorporation, of ABM as a viable means of propelling research into the mechanisms of learning.
In D. Abrahamson (Organizer), U. Wilensky (Chair), & R. Lesh (Discussant), Learning Complexity: Agent-based modeling supporting education research on student cognition in social contexts . Paper presented at the annual meeting of the American Educational Research Association, Chicago, IL.
The A–W 2005 Paper: ABM as a Methodology in the Service of Theory-of-Learning Research
Figure 1. Six snapshots from successive experimental runs in the “I’m Game!” model. The “I’m Game!” model (see Figure 1, above), built in the NetLogo agent-based modeling environment (Wilensky, 1999), implements sketches of “Piagetian” and “Vygotskiian” interpretations of human learning into a single model and examines learner–agents’ performance under “Piagetian,” “Vygotskiian,” and combined “Piagetian–Vygotskiian” conditions. “Players” (circled arrows) stand in a row (Figure 1a). They each roll a marble toward a target line. Some players undershoot the line, some overshoot it (Figure 1b). Players collect their marbles, adjust the force of their roll—based either on their own performance (“Piagetian”) or on their neighbors’ (“Vygotskiian”)—and, on a subsequent trial (Figure 1c), improve on their first trial—they have “learned” as individuals. Gradually, the group converges on the target line (see Figure 1d-1f, for three later attempts).
Figure 2. Interface of the “I’m Game!” NetLogo model, with results from multiple runs. We have run the model under a wide range of conditions so as to evaluate its capacity to simulate reliably core features of the embedded theoretical model. For example, Figure 2 (above) shows results from running the simulation within a particular parameter space (see the sliders and switches) under the three experimental conditions and a fourth, control condition (“random”). Group mean performance (distance from target) was ranked as “Piagetian–Vygotskiian” (nearest, so best), “Piagetian,” “Vygotskiian,” and “Random” (furthest, so worst). We thus express one possible interpretation of the complementarity of the Piagetian and Vygotskiian perspectives (e.g., Cole & Wertsch, 1996).
Proof of Existence: LCHC Responds to CCL’s Call for ABM-Based Theory-of-Learning Collaborative Research
As part of an effort to study distributed learning, members of the Laboratory of Comparative Human Cognition (LCHC) at the University of California, San Diego began using NetLogo as a modeling environment for learning and examined the Abrahamson–Wilensky (A–W) model of Piagetian and Vygotskiian learning. LCHC draws heavily on cultural historical activity theory developed by Vygotsky and others, and so they were especially interested in the expression of Vygotskiian learning embedded in the A–W model. The A–W model was introduced at a weekly laboratory meeting in several ways. First, the model was shown to the whole group while it ran through a series of simulations of learning. The parameters made available through the interface were explored. The underlying code was examined, but the amount of code and the relative unfamiliarity with NetLogo code by most of the Lab limited the utility of this examination. A printout of a subset of the code was distributed to the Lab members, with the core code
expressing the theory highlighted. Even though members of LCHC by and large did not know how to interpret NetLogo statements, a textual explanation of the core code was given by two members of the Lab who were familiar with NetLogo.
During the discussion, it became clear that the A–W implementation of the concept of “Zone of Proximal Development” was “simplex” – that is, learning in the ZPD embodied in the model depended on changes by the less skilled member of a pair of learners without any changes by the more skilled member. Several members of LCHC pointed out how the dynamic construction of a ZPD is “duplex,” that it involves both the learner and the “teacher” (the more skilled person). It occurred to Jim Levin that a relatively simple change to the model would implement a way in which teachers change their behavior depending on their knowledge of the level at which their students are performing. In this way, the members of LCHC were accepting the challenge in the Abrahamson & Wilensky (2005) paper describing this model: “We would welcome a critique that uses the existing model as a basis for expressing these constructs” (p. 28).
Figure 3: Interface of the “I'm Game!” A-W model as modified by Jim Levin to improve the
Vygotskiian learning.
Levin implemented a modification to the A–W model in which the better performing player modified their next move to be within the ZPD of the less well performing play, making a play
that was worse than they knew how to make, in order to help the less well performing player learn to play better. The modification is called the “-T” version (“T” for “teacher”), and a screenshot of the model after about 50 runs is shown in Figure 3, above.
While members of LCHC who viewed this modified model at a later meeting then raised further concerns about the model, especially about the potential for oversimplication with this simulation, they were also fairly impressed with the ease of modification. For researchers who are new to modeling-based inquiry, it is clearly easier to modify an existing model than to construct one.1 And it is even easier to explore the parameter space of an existing model.
A close examination of Figures 2 and 3 will reveal that not only does Figure 3 have two additional learning strategies but it also has different parameter values for three of the four parameters that take a range of numeric values, “ZPD,” “error,” and “#-Vygotskiian-neighbors”. If we set those three parameters to have the same values as in Figure 2, we see quite different results, as shown in Figure 4, below.
Figure 4: Interface of Levin’s modified A–W model, with a different set of numeric parameters.
1 NetLogo was designed so as to enable even children to be able to easily construct their own models. In the pedagogy developed at the CCL, we have found that an important step in the process of learning to construct one’s own model is to modify an existing model.
Note that in this run of the modified A–W model, the Piagetian strategy does well, but not as well as the combined P-V or P-V-T strategies. The Vygotskiian strategy does less well, and the Vygotskiian-t version does worse than the “Random” strategy. So it matters quite a bit which set of parameters are chosen in comparing these learning strategies. The Abrahamson & Wilensky paper (2005) explores some of the multi-dimensional parameter space, using NetLogo’s BehaviorSpace research tools (Wilensky, 2001; Tisue & Wilensky, 2004). But it is not surprising that the figures in that paper, presented to the Jean Piaget Society, show the model operating with parameters in which the “Piagetian” strategy does well, while the parameters shown in Figure 3 are chosen for the Levin & Cole (2007) paper which is part of this symposium.
So an important lesson is that parameterized models such as these are open to many different interpretations, depending on which regions of the parameter space you elect to explore. Further enhancements to ABM environments could facilitate this exploration by offering users interactive tools for navigating in the conceptual space expressed in the combinatorics of the parameters selected for modeling (for a very first sketch of possible development directions, see Appendix A for the entire code of the “-T” model, with print-based orientation coding to help constructionist collaborators navigate through procedures).
Conclusion
We have presented a case study—a proof of existence—of an agent-based model that enabled constructionist collaboration of two research groups in the refinement of a theoretical model. The following attributes of agent-based modeling enabled and supported the collaborative research. Agent-based modeling:
https://www.sodocs.net/doc/92867899.html,pels clear articulation of conceptual models and, inter alia, catalyzes personal
refinement of these models (Papert’s “debugging principle,”1980);
b.clearly circumscribes the theoretically known—the unknowns are expressed in the form
of “patchy” procedures that hard-code emergence into the simulation;
c.introduces modularity into the theoretical model, i.e., the program’s procedures isolate
each of the modeler’s conceptual assumptions. Thus, reviewers can readily locate each of the assumptions and implement modifications to the modeling code such that it agree
with their own theoretical assumptions;
d.constitutes a lingua franca that transcends and cuts across the ever-ramifying niches of
academic expertise (Jacobson & Wilensky, 2006);
e.stimulates and facilitates collaboration, including focused critique and modification
(Levin & Cole’s “distributed research” principle, this symposium). In particular, ABM
invites researchers who have not collaborated before to partake in a distributed
constructionist project.
Finally, the exteriorization of the model as a conceptual object with a felt agency of its own (“It’s alive!”) is perhaps more conducive to collaboration than are idiosyncratic models. Thus, ABM invites co-construction but not at the cost of constructive criticism.
Our point of departure was that learning is a complex phenomenon and is therefore suitable for study through agent-based modeling. And yet, the emergence of a modeling lingua franca for
studying learning can itself be studied through the lenses of complexity studies. In particular, ABM is an inherently interdisciplinary form of inquiry, because the meta-principles of diverse phenomena can be articulated from the perspective of numerous agents acting locally with more-or-less common rules (Jacobson & Wilensky, 2006). Therefore, we see a game-theoretical advantage for multiple researchers within and between disciplines to share a mode of inquiry. Collaborative research, a form of learning, is itself complex, and we perceive agent-based modeling as advantageous to all researcher–agents.
The Illusion of Agency: On What Agent-Based Modeling Is and What It Is Not
In closing, we wish to submit the following remark that may help in further clarifying our perspective on the nature of agent-based modeling vis-à-vis more traditional forms, especially with regards to its standing as embodying a form of inquiry. This remark should be taken as preemptive—it responds to anecdotal, rather than well-articulated, sentiments toward ABM that we have encountered as we share our work.
From the view-point of pre-ABM, Graham Cairns-Smith, a chemist focusing on evolution and consciousness, for whom modeling is an essential modus operandi, supposes that any educated consumer of models can distinguish between what a model is and what it is not:
Good analogies are like in some respects that are clearly understood; and unlike in other respects that are clear also. The organic chemist’s models, for example, are thought to
correspond to real molecules with respect to distances between atoms and so on, but not literally in all respects, and usually the distinctions are clear enough. Maybe the little
sticks are made of steel and they go rusty, but there is no danger at all of this being taken as a blinding new insight about the nature of molecules….Of course you should never
believe models, they only have an ‘as if’ status and they are likely to let you down any
time” (Cairns-Smith, 1996, pp. 46-47).
Any agent-based model is primarily no more and no less than a model. Granted, ABM incorporates avatars that often bear greater iconic resemblance to their phenomenal correlates than do static diagrams; agents are dynamic and interactive such that they induce a compelling anthropomorphic sense of agency; and agents operate fast, in parallel, and iteratively such that an uninformed model user is likely overwhelmed by the magnitude of perceptual information—it is as though the agent-based model takes on a life of its own, as though it is not an ‘as-if’ but ‘the thing itself.’ Indeed, we have observed many middle-school students cheering as a green triangle and a blue triangle compete against each other in a stochastic race across the computer interface. Students’ suspension of disbelief, projection of self, and syntonic embodiment of the stark geometrical forms appears vastly facilitated by the self-propelled motion (see Papert, 1980, on syntonicity). It needn’t take a Tamagochi reality pet to induce reality—blobs moving across the screen can do it, too. Thus, the na?ve phenomenology, if not the critical rationalization, of viewing a “run” of an agent-based model appears to share much in common with viewing that which is modeled therein.
However, no matter how great the verisimilitude, the complexity, and unpredictability of a model; however transfixing its dynamism; however enticing, engaging, and immersive its syntonic narrative…it is still a model. Indeed, one important reason that ABMs have proved so
powerful for learning is this verisimilitude. But this verisimilitude can have a cost of making people forget that they are working with a model, essentially the same kind of beast as an equation. Just as a quadratic equation can model ballistic motion but would not ever be expected to describe the texture of the projectile, so one should not expect an ABM to model the entire richness of a social interaction, only the aspects chosen to be modeled. As such, agent-based models—like any other model employed in the sciences and social sciences—is an expression only of carefully selected aspects of a phenomenon—aspects that are assumed to operate individually and in concert so as to factor into the key phenomenon of interest. Beyond that, appearances are meant only ostentatiously—they tag the avatar as ‘a model of X’—and this ostentation may cause confusion or frustration when the avatar does not deliver the richness of what X does in reality. This beguiling property of models is to be understood by the model user. The modeler certainly does not purport to offer a virtual animistic fetish—a Golem or Gollum—it is the viewer, and not the modeler, who invests the avatar with life and so expects this life to live to the brim. We hope this clarification will assist in fostering a critical yet productive use of agent-based modeling in collaborative scientific discourse.
References
Abrahamson, D., & Wilensky, U. (2005). Piaget? Vygotsky? I'm game!: Agent-based modeling for psychology research. Paper presented at the annual meeting of the Jean Piaget
Society.
Cairns-Smith, A. G. (1996). Evolving the mind: On the nature of matter and the origin of consciousness. New York: Cambridge University Press.
Cole, M., & Wertsch, J. V. (1996). Beyond the individual-social antinomy in discussions of Piaget and Vygotsky. Human Development, 39(5), 250-256.
Jacobson, M. J., & Wilensky, U. (2006). Complex systems in education: Scientific and educational importance and research challenges for the learning sciences. Journal of the Learning Sciences, 15(1), 11-34.
Levin, J. A., & Cole, M. (2007). Simulations as mediators for distributed research activity. Paper presented at the American Educational Research Association meetings, Chicago, IL. Papert, S. (1980). Mindstorms: Children, computers, and powerful ideas. NY: Basic Books. Papert, S. (1991). Situating constructionism. In I. Harel & S. Papert (Eds.), Constructionism (pp.
1-12). Norwood, NJ: Ablex Publishing Corp.
Smith, E. R., & Conrey, F. C. (2007). Agent-based modeling: A new approachy for theory building in social psychology. Personality and social psychology review, 11, 87-104. Tisue, S., & Wilensky, U. (2004, October). NetLogo: Design and implementation of a multi-agent modeling environment. Paper presented at the Agent 2004 conference, Chicago, IL. Wilensky, U. (2001). BehaviorSpace [Computer Software]. Evanston, IL: Center for Connected Learning and Computer Based Modeling, Northwestern University.
https://www.sodocs.net/doc/92867899.html,/netlogo/behaviorspace.
Wilensky, U. (1999). NetLogo. [Computer Software]. Evanston, IL: The Center for Connected Learning and Computer-Based Modeling, Northwestern University.
https://www.sodocs.net/doc/92867899.html,/netlogo/.
Appendix A – NetLogo Procedures of the “I’m Game” Model “-T,” With Proposed Print-Based Orientation System Supporting Constructionist Collaboration
;; Coded for core vs. infrastructure
;; Core theory content coded for high confidence, normal confidence, and less confidence
;; Modifications to the model by Dor Abrahamson & Uri Wilensky added by J A Levin 5 Feb 06
;; Further mods added by J A Levin 27 Apr 06
;; Original model available at:
;; https://www.sodocs.net/doc/92867899.html,/research/conferences/JPS2005/JPS2005.nlogo
;;This modification is available at:
;; https://www.sodocs.net/doc/92867899.html,/~jlevin/JPS2005-t.nlogo
globals [ target ;; the target line in the middle
max-dist ;; the maximum distance to that wall
runcount ;; the number of times the ball has been thrown since the start
ticks-left ;; the number of runs left of the current strategy
pavg vavg pvavg ravg vavgt pvavgt ;; lists of each run's distance halfway through ticks-left
;; ex. pavg = [ 2 3 17 ... ]
;; for comparison
current-strategy
result
]
turtles-own [ max-moves ;; the max-moves of the current throw
best-max-moves ;; the best throw of the agent (or, if no memory, the current throw)
best-max-moves-private ;; best throw, even if current throw different for teaching reasons (JAL)
moves-left ;; the number of moves left until the ball comes to rest
score ;; the current score (how far from target... lower is better)
best-score ;; best score (or, if no memory, current throw)
]
to setup
ca
set runcount 0
set-default-shape turtles "circle-arrow"
crt number-of-players
ask turtles [ set color color + 0.1 ] ;; to make trails a little easier
set pavg [0] set vavg [0] set pvavg [0] set ravg [0] set vavgt [0] set pvavgt [0]
;; initialized with zeros so i can use mean from the start
rerun
end
to rerun
cp
set runcount runcount + 1
;; random strategies lets you keep the simulation running picking from the available strategies
;; at random
if ( randomize-strategy-on-setup? ) [
let new-strategy random 6
if ( new-strategy = 3 ) [ set strategy "Random" set current-strategy 3]
if ( new-strategy = 2 ) [ set strategy "Piagetian" set current-strategy 2]
if ( new-strategy = 1 ) [ set strategy "Vygotskiian" set current-strategy 1]
if ( new-strategy = 0 ) [ set strategy "P-V" set current-strategy 0]
if ( new-strategy = 4 ) [ set strategy "Vygotskiian-T" set current-strategy 4]
if ( new-strategy = 5 ) [ set strategy "P-V-T" set current-strategy 5]
]
set ticks-left 30
; set ticks-left 16 ;; we use this one to stop each run half-way through. (see graph) This gives the info we need ;; , because the histogram represents values from half way through the run. For presentation, though, use "30"
set max-dist ( world-width )
setup-plots
setup-target
setup-turtles
display
end
to setup-plots
set-current-plot "avg distance"
set-plot-y-range 0 max-pxcor ;; most will fall within this range (all after a couple of steps)
set-current-plot-pen strategy ppu plotxy -1 0 ppd
set-current-plot "strategy avgs"
if length ravg > 1 [set-current-plot-pen "Random" plot-pen-reset plotxy 0 precision mean butlast ravg 2]
if length pavg > 1 [set-current-plot-pen "Piagetian" plot-pen-reset plotxy 1 precision mean butlast pavg 2]
if length vavg > 1 [set-current-plot-pen "Vygotskiian" plot-pen-reset plotxy 2 precision mean butlast vavg 2]
if length pvavg > 1 [set-current-plot-pen "P-V" plot-pen-reset plotxy 3 precision mean butlast pvavg 2]
if length vavgt > 1 [set-current-plot-pen "Vygotskiian-T" plot-pen-reset plotxy 4 precision mean butlast vavgt 2] if length pvavgt > 1 [set-current-plot-pen "P-V-T" plot-pen-reset plotxy 5 precision mean butlast pvavgt 2]
end
to-report get-strategy-color
ifelse ( strategy = "Random" )
[ report green ]
[ ifelse ( strategy = "Piagetian" )
[ report blue ]
[ ifelse ( strategy = "Vygotskiian" )
[ report red ]
[ ifelse ( strategy = "Vygotskiian-T" )
[ report magenta ]
[ ifelse ( strategy = "P-V-T" )
[ report cyan ]
[ report grey ]
]
]
]
]
end
to setup-target
;; the target is the line in the center of the screen, 2 patches thick
set target patches with [ abs ( pxcor ) < 1 ]
ask target [ set pcolor get-strategy-color ]
end
to setup-turtles
ask turtles
[
;; spread out the turtles evenly
set size ( world-height / count turtles )
setxy min-pxcor ((who * world-height / count turtles)) + 2
set score 100000
set best-score score
set heading 90
;; their max-moves are randomly distributed over the length of the playing field
set max-moves random max-dist
]
end
to go
;; the score is their distance from the target
ask turtles [ set score evaluate-score ] ;; the score is the distance of the turtle from the target line (evaluate-score is a reporter that provides this distance)
ask turtles [ adjust ] ;; act according to the strategy, e.g,. in Vygotskiian, you compare your scores to a neighbor and possibly update your score
reset
;; move all the turtles forward to their max-moves spot.
;; we can't just say "fd max-moves" because we want them to bounce off the wall
;; + leave a dissipating trail
ask turtles [ set moves-left limit-legal-distance max-moves ]
let moving-turtles turtles with [ moves-left > 0 ]
while [ any? moving-turtles ] [
set moving-turtles turtles with [ moves-left > 0 ]
ask moving-turtles [
move-x
if (trails?) [ set pcolor (color - 5) + ( 10 * ( max-moves - moves-left ) / max-moves ) ]
]
]
do-plots
;;if sum (sentence length pavg length vavg length pvavg length ravg) >= 5 [stop]
if (ticks-left < 0) and limited-run? [ display stop ]
if (ticks-left < 0) and not limited-run? [ display rerun
if ( strategy = "Random" ) [ if length ravg > 1 [ set result precision mean butlast ravg 2 ] ]
if ( strategy = "Piagetian" ) [ if length pavg > 1 [ set result precision mean butlast pavg 2 ] ]
if ( strategy = "Vygotskiian" ) [ if length vavg > 1 [ set result precision mean butlast vavg 2 ] ]
if ( strategy = "P-V" ) [ if length pvavg > 1 [ set result precision mean butlast pvavg 2 ] ]
if ( strategy = "Vygotskiian-T" ) [ if length vavgt > 1 [ set result precision mean butlast vavgt 2 ] ]
if ( strategy = "P-V-T" ) [ if length pvavgt > 1 [ set result precision mean butlast pvavgt 2 ] ]
]
set ticks-left ticks-left - 1
end
to move-x
set moves-left moves-left - 1
fd 1
if ( pxcor >= (max-pxcor - 1) )
[ set heading 270 fd 2 ]
end
to-report evaluate-score
report ( distancexy-nowrap 0 ycor ) ; the target has x-coordinate of 0, so this gives the horizontal distance of this turtle to the target
end
to-report limit-legal-distance [ val ]
report ( min ( list ( max-dist - 1 ) max ( list 0 val ) ) )
end
to do-plots
set-current-plot "avg distance"
let curr-mean mean values-from turtles [ evaluate-score ]
plot curr-mean
;; we sample after 15 steps. later than that, we lose information
if (ticks-left = 1) [
if strategy = "Piagetian" [ set pavg fput round curr-mean pavg ]
if strategy = "Vygotskiian" [ set vavg fput round curr-mean vavg ]
if strategy = "P-V" [ set pvavg fput round curr-mean pvavg ]
if strategy = "Random" [ set ravg fput round curr-mean ravg ]
if strategy = "Vygotskiian-T" [ set vavgt fput round curr-mean vavgt ]
if strategy = "P-V-T" [ set pvavgt fput round curr-mean pvavgt ]
]
end
to reset
cp
setup-target
ask turtles [ set heading 90 set xcor min-pxcor ]
display
end
to adjust
if strategy = "Random" [
r-adjust
set max-moves best-max-moves
stop
]
if strategy = "Piagetian" [ p-adjust ]
if strategy = "Vygotskiian" [ v-adjust ]
if strategy = "Vygotskiian-T" [ v-adjust-t ]
if strategy = "P-V" [ pv-adjust ]
if strategy = "P-V-T" [ pv-adjust-t ]
if (strategy = "Vygotskiian") or (strategy = "Vygotskiian-T") ;; or (strategy = "P-V-T")
[
set max-moves limit-legal-distance
( best-max-moves + ( random-normal 0 ( error * best-score / max-dist ) ) )
stop
]
ifelse ( xcor > 0 ) [
set max-moves limit-legal-distance
( best-max-moves + (- abs random-normal 0 ( error * best-score / max-dist ) ) )
] [
set max-moves limit-legal-distance
( best-max-moves + (abs random-normal 0 ( error * best-score / max-dist ) ) )
]
end
to p-adjust
;; if your score is better, that's your new best, otherwise stick with the old
if (score < best-score) [
set best-score score
set best-max-moves max-moves
]
end
to v-adjust-t ;; modified from v-adjust by JAL
if (abs((max-pxcor / 2) - best-max-moves) > abs((max-pxcor / 2) - best-max-moves-private)) [set best-max-moves best-max-moves-private] ;; restore best knowledge
let fellow nobody
while [ fellow = nobody or fellow = self ] [ set fellow turtle (who + ( - (#-Vygotskiian-neighbors / 2) ) + random (1 + #-Vygotskiian-neighbors ) ) ]
;; look randomly to one of your neighbors
ifelse (best-score > best-score-of fellow) and (best-score - ZPD <= best-score-of fellow) ;; if other is better and within ZPD
[
set best-score best-score-of fellow
set best-max-moves best-max-moves-of fellow
]
[
set best-score score
set best-max-moves max-moves
;; below is the new code that attempts to model the "teacher's" awareness of the "student" (JAL)
set best-max-moves-private 0
if (best-score < best-score-of fellow) ;; better than other (ie. closer to the target)
[
set best-max-moves-private best-max-moves ;; save own knowledge of own best move, since we're going to move suboptimally for teaching purposes; assumes perfect memory, which is of course questionable (JAL)
ifelse (best-max-moves < best-max-moves-of fellow)
[
set best-max-moves max (list best-max-moves ((best-max-moves-of fellow) - ZPD )) ;; target is to the left
]
[
set best-max-moves min (list best-max-moves ((best-max-moves-of fellow) + ZPD )) ;; target is to the right
]
]
]
end
to v-adjust ;; original Abrahamson & Wilensky v-adjust (JAL)
let fellow nobody
while [ fellow = nobody or fellow = self ] [ set fellow turtle (who + ( - (#-Vygotskiian-neighbors / 2) ) + random (1 + #-Vygotskiian-neighbors ) ) ]
;; look randomly to one of your neighbors
;; if the score is better and it is within your ZPD, use their max-moves.
ifelse (best-score > best-score-of fellow) and (best-score - ZPD <= best-score-of fellow)
[
set best-score best-score-of fellow
set best-max-moves best-max-moves-of fellow
]
[
set best-score score
set best-max-moves max-moves
]
end
to pv-adjust-t ;; modified pv-adjust with added Vygotskian "teacher" effect (JAL)
if (abs((max-pxcor / 2) - best-max-moves) > abs((max-pxcor / 2) - best-max-moves-private)) [set best-max-moves best-max-moves-private] ;; restore best knowledge (JAL)
let fellow nobody
while [ fellow = nobody or fellow = self ] [ set fellow turtle (who + ( - (#-Vygotskiian-neighbors / 2) ) + random (1 + #-Vygotskiian-neighbors ) ) ]
;; look randomly to one of your neighbors
;; maximize your own score and...
if ( score < best-score )
[
set best-score score
set best-max-moves max-moves
]
;; check it against your neighbor's score
ifelse (best-score > best-score-of fellow) and (best-score - ZPD <= best-score-of fellow)
[
set best-score best-score-of fellow
set best-max-moves best-max-moves-of fellow
]
[
set best-score score ;; present in v-adjust but not in pv-adjust? (JAL)
set best-max-moves max-moves ;; present in v-adjust but not in pv-adjust? (JAL)
;; below is the new code that attempts to model the "teacher's" awareness of the "student" (JAL)
set best-max-moves-private 0
if (best-score < best-score-of fellow) ;; better than other (ie. closer to the target)
[
set best-max-moves-private best-max-moves ;; save own knowledge of own best move, since we're going to move suboptimally for teaching purposes, assumes perfect memory (JAL)
ifelse (best-max-moves < best-max-moves-of fellow)
[
set best-max-moves max (list best-max-moves ((best-max-moves-of fellow) - ZPD )) ; target is to the left
]
[
set best-max-moves min (list best-max-moves ((best-max-moves-of fellow) + ZPD )) ; target is to the right
]
]
]
end
to pv-adjust ;; original pv-adjust from Abrahamson & Wilensky (JAL)
let fellow nobody
while [ fellow = nobody or fellow = self ] [ set fellow turtle (who + ( - (#-Vygotskiian-neighbors / 2) ) + random (1 + #-Vygotskiian-neighbors ) ) ]
;; look randomly to one of your neighbors
;; maximize your own score and...
if ( score < best-score )
[
set best-score score
set best-max-moves max-moves
]
;; check it against your neighbor's score
if (best-score > best-score-of fellow) and (best-score - ZPD <= best-score-of fellow)
[
set best-score best-score-of fellow
set best-max-moves best-max-moves-of fellow
]
end
to r-adjust
;; random strategy changes max-moves to a random number x if it's not at the wall
;; where 0 < x < max-dist
;; if it is at the target, it stops changing.
ifelse ( (abs pxcor) > 0 )
[
set best-max-moves ( random max-dist )
] [
set best-max-moves ( max-dist / 2 ) - 1
]
end
As的用法
As的用法 as...as 的用法比较多，中间除了加形容词、副词，名词也可以的（请看下面第2点）： as...as的结构： as + 形容词或副词原级+ as 1）在否定句或疑问句中可用so… as。 He cannot run so/as fast as you. 2）当as… as 中间有名词时采用以下格式。 as +形容词+ a +单数名词 as + many/much +名词 This is as good an example as the other is. I can carry as much paper as you can.. 3）用表示倍数的词或其他程度副词做修饰语时，放在as的前面。 This room is twice as big as that one. Your room is the same size as mine. 4) 倍数+ as + adj. + as<=> 倍数+ then + of This bridge is three times as long as that one. This bridge is three times the length of that one. Your room is twice as large as mine. Your room is twice the size of mine. the population of our students was _that of theirs. A twice as large as B twice as much as A twice as large as 是对的。 这是倍数/分数表达的词序问题。“A 是B的几倍/几分之一。。。”的正确词序是：A+be 倍数/分数+as +形容词/+as B 此外，与population搭配的形容词是large/great/big，如：a big/large/great population，或者the population is big/large/great,而不能用much或者many。
as的几种固定用法
as的几种固定用法 今天和大家一起学习一下at用法，快来一起学习吧，下面就和大家分享，来欣赏一下吧。 语法必看：as的几种固定用法 在英语中as是很常见的一个小词，我们最熟悉的是它作为介词的相关用法，但as还可以充当副词和连词等词性，并构成一些固定搭配，今天准备跟同学们分享一下as的几种基本用法，赶紧收藏学习吧。 as作介词时，一般有两种含义： “如，像”; They got united as one man.他们团结得像一个人一样。 “充当，作为”; As a painter，she was famous.作为画家，她很出名。 as作连词时，常用来连接主句和状语从句 引导时间状语从句，含义为“当...的时候”;
I was startled as he opened the door.他一开门，我吓了一跳。引导原因状语从句，含义为“因为，由于”，与because的用法相近;I must stop walking now，as I have a headache.我必须停下来，因为我头很疼。 引导让步状语从句，含义为“虽然，尽管”; Strange as it may seem，it is true.尽管这事看上去很奇怪，但却是真的。 as还用于一些固定搭配之中： as far as I am concerned 在我看来 as soon as 一...就... as far as I know 据我所知 除上述用法外，as还可用作连词引导状语从句。As作为引导词可连接多类状语从句，如时间状语从句、原因状语从句、方式状语从句或比较状语从句。因此，还望同学们能够记住as的各种用法，一些含as的固定短语还可以用在写作当中。 【基础语法】介词on at用法大全 ?on常见用法 1.动词+on
as的用法总结
as可用作连词，引导比较、时间、原因及方式等四种状语从句，应注意的是，引导的比较状语从句往往有省略；引导的时间状语从句一般用一般现在时而不用一般将来时；引导原因状语从句时与“Because”和“since”引导的从句比较起起来语气最弱。 as作介词，意思时“作为”，“以……身份”。例如：He came to China as a tourist five years ago. 而表示像…一样时，like通常作介词用，而介词后面通常接名词，代名词和动名词。 as in Australia“和在澳大利亚一样”。 as……as AS +adj(原级)+AS AS +adv(原级)+AS as soon as 一……就 as soon as possible 尽可能快地 as early as possible 尽可能早的 as carefully as you can 尽可能认真地 as careful as you can 尽可能认真的 1。as是连词，表示“随着”的意义。lives是life的复数，意义是“生活”。
2。as是连词，表示“因为，由于”的意义。wanting是现在分词，表示主动意义，wanting to buy cars=who want to buy cars。 As的用法小结 as可以作连词、介词及关系代词和副词。现将其用法小结如下: （一）、as作连词的用法: 1.作“在-------期间，当----的时候”引导时间状语从句.注意与when、while的用法区别。 ①下列情形时,只用as, 而不用when或while。 1)用于表示同一个人的两种动作交替进行,指一边----一边. 如: The girl sings as she goes to school. He looked behind occasionally as he went forward in the forest. 2)表示两个同步发生的动作或行?意思是随着-----的发展.如: As time went on / by, she became more and more beautiful . As children get older, they become more and more interested in everything. 3)表示两个短暂行为或事情几乎同时发生.如:I watched her as she read the book. I thought of it just as you opened your mouth. Just as the flying worm hit her face, she gave a loud cry. 4)接在名词后面表示某一个年龄段时.如: As a young man, he was active in sports.
英语中as用法详解
as 用法详解 1.用作连词，表示让步(意为：虽然，尽管)，要用于倒装句，且倒装后位于句首的名词通常不用冠词(等于though, 但语气稍弱)。如： Child as[though]he was, he did quite well. 他虽是个孩子，但已干得很不错。 2.用作关系代词，主要用法有二。如： (1)用在such, same, as等之后，引导限制性定语从句。如： This is the same watch as I lost. 这块表跟我丢失的那块一样。 Such men as (=Those men who) heard of him praised him. 听说过他的人都赞扬他。 (2)单独用作关系代词，引导非限制性定语从句，可放在主句之前(常译为：正如)或之后(常译为：这一点)，且主从句之间一般要用逗号隔开。如： He was absent, as is often the case. 他缺席了，这是常有的事。 As was expected, he succeeded at last. 正如我们所料，他终于成功了。 3.用来表示目的，下面两种句子结构都可以。如： 仔细写以便把每句话都写清楚。 正：Write carefully so as to make every sentence clear. 正：Write so carefully as to make every sentence clear. 4.在正式文体(尤其是文学体裁)中，as后可用倒装语序表示“…也一样”这类意思(现代英语通常so表示这一用法)。如： He travelled agreat deal, as did most of his friends. 他去过许多地方旅游，他的多数朋友也是一样。 He plays the piano, as does his mother. 他会弹钢琴，他母亲也会。 注：用作连词，表示原因(参见because)，表示时间(参见when)。 5.用于as if as though, 意为“仿佛”、“似乎”。两者一般可通用(但注意不能说as although)。用法上注意几点： (1)as if [though] 可引导状语从句和表语从句，从句谓语根据语义的要求，可用陈述语气(若可能为事实)或虚拟语气(若为假设或不太可能是事实)。如： It looks as if it is going to rain. 天似乎要下雨了。 It seems as if you’re right. 似乎你是对的。 He talks asthough he knew everything. 他夸夸其谈，好像无所不知。 When apencil is partly in a glass of water, it looks as if it werebroken. 当把一支铅笔的一部分放在一杯水里时，它看起来好像折断了似的。 当主句谓语是过去时态时，从句谓语常可用陈述语气。如： He pausedas though he found a difficulty. 他停了下来，似乎遇到了一个难题。 She felt as if she lost something. 他觉得似乎丢了什么东西。 (2)当从句主语与主句主语一致，且从句谓语中又包括有动词be时，从句主语及其谓语中的动词be通常可以省去。如：
as用法
As的用法 一）as作副词，表示程度，意为“同样地”。在“as．．．as．．．”，“not as．．．as．．．”结构中的第一个as是副词，作“和／与．．．(不)一样”解。 eg：Jack is as tall as his father．杰克和他的父亲一样高。 e.g. He doesn't speak English as／so fluently as you．他的英语说得不如你流利。注意：as…as可用于表示两个人或物不同性质的比较，表示程度相当或相等，意为“…而…”。 e.g. The prison are as over-crowed as the farmlands are empty. 监狱里人满为患，而土地则无人耕种。 e.g. He was as handsome as his wife was beautiful. 他长得非常英俊，他的妻子长得也非常漂亮。 as…as还可用于表示一个人或物不同性质的比较，表示程度相当或相等，意为“既…又…”。 e.g. She is as kind as honest。她既诚实又善良。 e.g. The problems are as numerous as (they are) trivial。问题又多又繁琐。 二）as作介词 1．作“如，像”解。 eg：They got united as one man．他们团结得像一个人一样。 2．作“充当，作为”解。 eg：As a writer，he was famous．作为作家，他是很有名的。 三）as作连词，常用来连接主句和状语从句。 1．引导时间状语从句，作“当．．．的时候”解，有“随着．．．”之意，与while意义相近，强调两个动作同时发生；或某事一发生，另一事立即发生。 eg：He shouted aloud as her ran along．他一边往前跑，一边高声地呼喊。 e.g. I was startled as he opened the door．他一开门，我吓了一跳。 as作连词，相当于when。 eg；As a little boy (When he was a little boy)he began to learn to play piano．他小时候就开始学弹钢琴。 2．引导原因状语从句，作“因为，由于”解，与because的用法相近。 eg；I must stop writing now，as I have rather a lot of work to do．我必须停笔了，因为我还有许多工作要做。 3．引导方式状语从句或比较状语从句，作“正如，(如)像”解。例 eg：As in your country，we grow wheat in the north and rice in the south．正如(像)你们国家一样，我们在北方种小麦，在南方种大米。(方式状语从句) e.g. When at Rome，do as Romans do．入乡随俗。(方式状语从句) “as…, so…”结构意为“正如…也…，犹如…一样”，表示的是比拟关系。As引导的是含有比拟意义的方式状语从句，so相当于“in the same way, in the same proportion”，引导主句。该结构有以下几个特点： 1）as前可加just，表示强调；2）as从句可以居前，也可以居后，在前时用逗号隔开，在后时不用；3）as从句居前时，主句常用倒装结构；4）“as…, so…”结构可用“what…, that…”
As的各种用途
As的各种用途 在英语里，“as”这个字神通广大，有各种用法，下面便是其中九种： 一、用作介词。 用作介词的as 1.表示“如，像”。例如：They got united as one man./She spoke of me as her dearest friend. 2. 表示“作为、当作”。例如：As a League member, you should think more of others. 3. 与某些动词搭配，表示“把……当作……”，如：look on…as…, regard…as…, treat…as…, consider…as…, think of…as…, see…as…等。其中consider…as… 中的as可以省略。as与famous或known搭配，表示“作为……而出名”。 例Linda worked for the Minnesota Manufacturing and Mining Company, ________ as 3M. A. knowing B. known C. being known D. to be known 解析：如果熟悉be known as这一短语，运用有关非谓语动词的常识，可选出正确答案B。 例如： （1）As the new manager of the company, Wilcox assured the staff that he would work for the benefits of the company. （2）The scenery here is beautiful as a picture. 二、用作表示“程度”的副词。例如： 修饰形容词或副词，表示程度，意为“同样地”。例如：He swims fast, but I swim just as fast. 但它通常构成表示比较的结构“as…as…”，“not as…as…”。此结构中第一个as是副词，第二个as是连词。否定结构中的副词as可以由so代替。as…as possible /one can也属于此用法。例如：It is generally believed that teaching is as much an art as it is a science.
as well as的用法
as well as的用法 曲靖市第二中学 as well as后接动词到底该用什么形式，这是一比较复杂的问题，归纳起来有以下几点值得注意： 一、as well as构成同级比较结构，意为“和…… 一样好”。 第一个as是副词，第二个as是连词，引导一个状语从句，表示同级比较。 在否定句中可用not so well as代替not as well as。 He speaks English as well as a native speaker. 他讲的英语是和英语是母语的人一样好。 He can operate the machine as well as I do. 他操作这台机器和我一样熟练。 He speaks English as well as she. 他说英语说得跟她一样好。 She plays every bit as well as the men. 她打得一点不比男人们差。 He doesn’t play half so well as his sister. 他演奏的水平不及他姐姐的一半。 He sings as well as, if not better than, Mary. 他要是唱歌不比玛丽唱得更好，但至少也是一样好。 She cooks as well as her mother (does). 她烧菜和妈妈一样好。 二、as well as用作连词，连接两个并列的同等成分，其意义为“不但……而且……”，“既……又……”，这时相当于not only ... but also ...。 它所连接的部分既可以是单词、短语，也可以是句子。 在A as well as B的结构里，语意的重点在A，不在B。翻译时要特别注意。The girl is lively as well as healthy.（连接两个表语）这女孩既健康又漂亮。 He can speak Spanish as well as English. (连接两个宾语) 他不但会说英语，而且会讲西班牙语。 In China, as well as in Canada, the weather changes from season to season. (连接两个状语) 中国的天气和加拿大一样随季节的变化而变化。 The teacher as well as the students enjoys listening to English songs.（连接两个主语）老师和学生都爱听英语歌曲。 He grows flowers as well as vegetables. 他既种菜也种花。 She shares (in) my troubles as well as my joys. 她与我同甘共苦。 They have a flat in town as well as in the Country. 他们在城里有一套公寓，在乡村还有一所房子。 We are repairing the roof, as well as painting the walls. 我们既油漆墙壁，又修房顶。 It is important for you as well as for me. 这对我很重要，对你也很重要。 Lily as well as her parents is very fond of classical music. 不但莉莉的父母，而且连
As的详细用法
As的详细用法 As 竟然有这么多用法，这个意思你绝对没想到!?快来一起学习吧。下面就和大家分享，来欣赏一下吧。 As 竟然有这么多用法，这个意思你绝对没想到! PART ONE “as … as … 像…一样…” ，这个结构是“as”最基本的用法之一。 eg: 1.You look as young as your daughter. 你看起来和你的女儿一样年轻。 2.I don’t think James is as nice as you think he is. 我不认为詹姆士像你想的那样善良。 3.The smartphone is as good as new. 这台智能手机的状态和新的一样。
PART TWO “as”表达“作为，以…的身份;如同”的意思。 eg: I used to work part-time as a barista. 我曾经兼职做过咖啡厅服务员。 Although Tony has a crush on Jess, she sees him as just a friend. 虽然托尼非常迷恋杰丝，但杰丝只把他当作朋友而已。 This painting is now regarded as a masterpiece. 如今，这幅画被人看成是杰作。 03 PART THREE “as”可以用来说明事情的原因，译作“由于、因为”，它的语气要弱于表示直接因果关系的从属连词“because”。
Eating fruit and vegetables every day is essential, as they are healthy and nutritious. 每天吃水果蔬菜非常重要，因为这些食物既健康又有营养。 This is a fake pound coin, as the genuine one has different shape. 这枚一镑硬币是假的，因为真币的形状不同。 04 PART FOUR “as”作连词也有“随着…”或“在…的同时”的意思。 eg: We all jumped as the monster appeared on the screen. 当那只怪物出现在荧屏上时，我们都被吓了一跳。 The baby started to giggle as she heard the music. 这个婴儿听到音乐后开始咯咯地笑了起来。 05
as...as的用法
as...as的用法 as...as意为"和……一样"，表示同级的比较。使用时要注意第一个as为副词，第二个as 为连词。其基本结构为：as+ adj./ adv. +as。例如： (1)This film is as interesting as that one.这部电影和那部电影一样有趣。 (2)Your pen writes as smoothly as mine.你的钢笔书写起来和我的一样流畅。 其否定式为not as/so +adj./ adv. +as。例如： This dictionary is not as/so useful as you think.这本字典不如你想象的那样有用。 若有修饰成分，如twice, three times, half, a quarter等，则须置于第一个as之前。例如：Your bag is twice as expensive as mine.你的袋子比我的贵一倍。 几个关于as...as的常见句型： (1)as...as possible Please answer my question as soon as possible.请尽快回答我的问题。 (2)as...as usual/before She looks as pretty as before.她看起来和以前一样漂亮。 (3)as long as... (引导条件状语从句) It took us as long as three years to carry out the plan.我们花了长达三年的时间才完成这项计划。 (4)as far as He walked as far as the railway station yesterday evening.昨天傍晚，他一直散步到火车站。 (5)as well as She cooks as well as her mother does.她烧菜烧得跟她母亲一样好。 一些带有as...as结构的常见短语归纳： as busy as a bee像蜜蜂一样忙碌as easy as ABC像ABC一样容易as deep as a well像井一样深 as light as a feather像羽毛一样轻as soft as butter像黄油一样软as rich as a Jew像犹太人一样富裕 一、“as(否定句中可用so)＋adj．／adv．＋as…”，其基本意思为“……和……一样”。但在实际应用中，此结构在不同的语境中含义差异较大。 1、表示不同人或物同一性质的比较，意为“……和……一样……"。 The tree is as tall as the building(is)．这棵树和那栋楼一样高。 Michael is as bright as George(is)．迈克尔和乔治一样聪明。 2、表示同一个人或物不同性质的比较，意为“既……又……” Alice works as happily as(she)plays(happily)．艾丽丝愉快地工作，尽情地玩。 Danny is not so wise as he is witty．丹尼为人风趣，但欠明智。 3、用于表示两个人或物不同性质的比较，表示程度相同或相当，意为“……而……”。 He was as handsome as his wife was beautiful．他长得非常英俊，他的妻子也长得非常漂亮。 He was as experienced as his brother was green．他经验丰富，而他兄弟却涉世未深。 二、as…as结构的另一种形式是“as much／many＋名词＋as＋从句”。 Mary has written as many essays as her brother．玛丽写的文章篇数和她弟弟一样多。 It is as much your fault as your wife’s．这既是你的过错，也是你妻子的过错。 Henry is as much a hypocrite as·John．亨利跟约翰一样是个伪君子。 三、若有修饰成分，如twice，three times，half，a quaer等，必须置于第一个as之前。 You are not half as clever as you think you are．你可不像自己想象的那么聪明。 She isn’t going out with a man who is twice as old as she． 她不打算和一个比她大一倍的人出去散步。 You’ve made just as many mistakes as I have．我和你犯的错误一样多。
英语语法中as的三大用法
英语语法中as的三大用法 第一个单词prospect，它表示的意思是“前景”，常用复数形式prospects。第二个单词wage, 是工资的意思，通常指的是按短期（小时，星期）计算的工资，要注意和salary 进行区分，salary指按长期（月，年）计算的工资。最后一个单词transform，这个词很好记，单词中“trans” 这个前缀就是“转变、变换”的意思，所以transform的意思就是转变、改变。
首先句子的主语非常好判断，那就是“parents and students”，后面跟的是who 引导的定语从句，从句中谓语动词是invested。后面的worry就是整个主句的谓语动词了。句子中的as, 连接的是前后两句话，但因为as的含义非常多，在此具体是什么意思，我们还要等到看完整个句子才能做出判断。 As后面的从句中，“technological advances and changes”, 是从句的主语，谓语动词为transform。通过分析句子含义，我们可以发现，as前面其实是事情发生的结果（父母和学生担心毕业前景），而as后面是事情发生的原因（国内和全球市场变化，薪资降低），所以as在这里的含义我们可以理解为“由于”。 经过拆分，现在句子已经非常清晰了。刚才我们讲到了as 的含义有很多，所以今天，艾伦英语部落就为你进行as这个词常用用法的拓展，一共有三种。
这里的as就是作为副词出现，同时句子中也用到了经常和as搭配使用的短语“as…as”, 表示“和…一样”。 我们看到第一个句子中，“作为一个中文老师”，as用作介词出现在这里，表达的意思是“作为”，第二个句子里， as 的意思为“如同，好像”，句子中提到的“像小男孩一样”。同样是介词，但在两个句子中表达的意思完全不同，大家要注意区分。 ③ 用作连接词，可以表示“当…时”，“因为”，“随着”等。 As作连接词使用时，含义比较多，我们分别来看几个例句：
as用法大全
一、as作连词的用法 1. as...as的用法 as...as意为"和……一样",表示同级的比较.使用时要注意第一个as为副词,第二个as为连词.其基本结构为：as+ adj./ adv. +as.例如： (1)This film is as interesting as that one.这部电影和那部电影一样有趣. (2)Your pen writes as smoothly as mine.你的钢笔书写起来和我的一样流畅. 其否定式为not as/so +adj./ adv. +as.例如： This dictionary is not as/so useful as you think.这本字典不如你想象的那样有用. 若有修饰成分,如twice, three times, half, a quarter等,则须置于第一个as之前.例如： Your bag is twice as expensive as mine.你的袋子比我的贵一倍. 几个关于as...as的常见句型： (1)as...as possible Please answer my question as soon as possible.请尽快回答我的问题. (2)as...as usual/before She looks as pretty as before.她看起来和以前一样漂亮. (3)as long as... (引导条件状语从句) It took us as long as three years to carry out the plan.我们花了长达三年的时间才完成这项计划. (4)as far as He walked as far as the railway station yesterday evening.昨天傍晚,他一直散步到火车站. (5)as well as She cooks as well as her mother does.她烧菜烧得跟她母亲一样好. 一些带有as...as结构的常见短语归纳： as busy as a bee像蜜蜂一样忙碌 as easy as ABC像ABC一样容易 as deep as a well像井一样深 as light as a feather像羽毛一样轻 as soft as butter像黄油一样软 as rich as a Jew像犹太人一样富裕
as...as...的用法总结
as...as...的用法总结 as...as...的用法其实很简单，快来一起学习吧。下面就和大家分享，来欣赏一下吧。 as...as...的用法其实很简单 as是英语中用得比较多的一个词，也是很多小伙伴反映较难掌握的一个词。as的词性较多，用法也较复杂，我们今天先讲一下很常用的as...as...的用法 as...as...的基本意思为“与……一样”，"像...一样的"。例如： The tree is as tall as the building(is). 这棵树和那栋楼一样高。 其中的第一个as为副词，其后通常接形容词或副词(用原级) ，第二个as可用作连词(后接从句)。 as...as...的用法 1 基本用法： 虽然都是as...as...但在不同的语境下，意思也不同。例如：
1.表示不同人或物同一性质的比较，意为“……和……一样……" Michael is as bright as George(is). 迈克尔和乔治一样聪明。 2. 表示同一个人或物不同性质的比较，意为“既……又……” Alice works as happily as(she)plays(happily). 艾丽丝愉快地工作，尽情地玩。 3. 用于表示两个人或物不同性质的比较，表示程度相同或相当，意为“……且/而……”。 He was as handsome as his wife was beautiful. 他长得非常英俊，他的妻子也长得非常漂亮。 He was as experienced as his brother was green. 他经验丰富，而他兄弟却涉世未深。 使用时应注意以下几点： 1. 在否定句中，第一个as 也可换成so： He doesn’t study as [so] hard as his brother. 他学习不如他弟弟努力。 2. 在该结构的两个as之间通常接形容词或副词的原级，但若涉及数量或程度，可用“as much+不可数名词+as”和“as many+可数名词复数+as” ：
as 用法详解
as 用法详解 1.用作连词，表示让步 (意为：虽然，尽管)，要用于倒装句，且倒装后位于句首的名词通常不用冠词 (等于though, 但语气稍弱)。如： Child as[though]he was, he did quite well. 他虽是个孩子，但已干得很不错。 2.用作关系代词，主要用法有二。如： (1)用在such, same, as等之后，引导限制性定语从句。如： This is thesame watch as I lost. 这块表跟我丢失的那块一样。 Such men as (=Those men who) heard of him praised him. 听说过他的人都赞扬他。 (2)单独用作关系代词，引导非限制性定语从句，可放在主句之前 (常译为：正如)或之后 (常译为：这一点)，且主从句之间一般要用逗号隔开。如： He was absent, as is often the case. 他缺席了，这是常有的事。 As was expected, he succeeded at last. 正如我们所料，他终于成功了。 3.用来表示目的，下面两种句子结构都可以。如： 仔细写以便把每句话都写清楚。 正：Write carefully so as to make every sentence clear. 正：Write so carefully as to make every sentence clear. 4.在正式文体 (尤其是文学体裁)中，as后可用倒装语序表示“…也一样”这类意思 (现代英语通常so表示这一用法)。如： He travelled agreat deal, as did most of his friends. 他去过许多地方旅游，他的多数朋友也是一样。 He plays the piano, as does his mother. 他会弹钢琴，他母亲也会。 注：用作连词，表示原因 (参见because)，表示时间 (参见when)。 5.用于 as if as though, 意为“仿佛”、“似乎”。两者一般可通用 (但注意不能说 as although)。用法上注意几点：
高考英语易错词汇解析asif用法举例辨析
as if用法举例辨析 It seems that / as if ...用法举例辨析 此句型实质上是“主+系+表”结构。其中it是人称代词，并无实意，指的是某种情况，seems为系动词，that / as if 引导表语从句。 一、It seems that …表示“看起来……”。 强调根据一定的事实所得出的一种接近于实际情况的判断，可以说表示事实。(可以转换成“名词或代词 + seem +动词不定式”句型，其意不变，如果动词不定式为to be + 形容词时，to be往往省略。) It seems that it is more difficult for women to get to the top of the company. 妇女似乎更难提升到公司的最高职位。 It seems that no one knows what has happened in the park. (No one seems to know what has happened in the park.) 似乎没有人知道在公园里发生了什么事。 It seems to me that he has known everything. (He seems to have known everything) 在我看来他似乎什么事都知道了。 It seems to me that Mr. White will not come again. 依我看，怀特先生不会再来了。 It seems that she is happy. 她似乎很高兴。 It seems that he likes his new job. 他看起来很喜欢他的新工作。 It seems that they don't like the idea. 他们似乎不喜欢这个主意。 It seems that he is lying.看来他在撒谎。 二、It seems as if ...也表示“看起来……，似乎……”。 使用该句型表示可能：若从句表示的是很可能实现的事情，从句谓语用陈述语气;若从句表示的是与事实相反的假设或实现的可能性很小的事情时，则从句谓语要用虚拟语气。 1、从句的谓语动词常用虚拟语气来表示所设想的事情不真实或不可能发生/存在。 It seems as if she had read this novel.看来她好像看过这本小说。 It seems as if it were autumn. 现在仿佛是秋天似的。 It seems as if it were spring already. 似乎春天已经来了。 It seems as if he knew everything. 好像他什么都知道似的。 It seems as if she had been to England. 好像她到过英国似的。 It seems as if they had never seen each other before. 仿佛他们以前从未见过面。It seems as if he were in a dream.看来他像是在做梦。 It seemed as if the good man were trying to teach us all he knew at this last lesson.这个好心人仿佛要在最后这一堂课上把他的全部知识教给我们。 2、如果可能性很大则用陈述语气，或者句中的情况可能发生或可能被设想为真实，则仍然用陈述语气。 It seems as if our team is going to win.看来我们队要胜了。 It seems as if he has learned by heart a speech written by someone else.看来他像是背了一篇由别人写的演讲稿。 It seems as if he has been at the scene of the crime似乎他好像在犯罪现场。 It seems as if it is going to rain. (It seems to rain) 看来天要下雨了。
“as…as”的用法
“as…as”的用法 “as + 形容词或副词原级+ as…” 意为“和……一样”，表示同级的比较，但要注意第一个as 为副词，第二个as 为连词。根据比较对象和性质的不同又可以把此种结构分为三种：（一）“ as … as …结构”用于表示不同人或物同一性质的比较，意为“……和……一样……" 。例如 The tree is as tall as the building(is) ． 这棵树和那栋楼一样高。 注意此结构也常用于形象比喻中：as cool as cucumber 沉着的，as stubborn as a mule 非常顽强的，倔强的，as clear as crystal 非常透明的，as poor as a church mouse 非常贫困的，as busy as a bee 像蜜蜂一样忙碌，as easy as ABC 像ABC 一样容易，as deep as a well 像井一样深，as light as a feather 像羽毛一样轻，as soft as butter 像黄油一样软。这些短语生动活泼，浅显易懂，记住这些短语可以提高学生的英语写作和口语能力。 （二）“ as … as …结构”用于表示同一个人或物不同性质的比较，意为“既……又……” 。例如： This swimming pool is as wide as it is long. 这个游泳池宽度和长度相等。 （三）“ as … as …结构”用于表示两个人或物不同性质的比较，表示程度相同或相当，意为“……而……”。在否定句中结构是“ so/as…as… ”。例如： He was as handsome as his wife was beautiful ． 他长得非常英俊，他的妻子也长得非常漂亮。 He was as experienced as his brother was green ． 他经验丰富，而他兄弟却涉世未深。 He cannot run so/as fast as you. 他没你跑得快。 Danny is not so/as wise as he is witty ． 丹尼为人风趣，但欠明智。 需要注意的是：“as + 形容词或副词原级+ as…” 结构的变体形式是“ as + 形容词+ a + 单数名词+ as 或as + many/much + 名词+ as ”。意为“……和……一样…… " 。例如：
As的用法归纳as的用法归纳
As 的用法归纳 as的用法很多，又比较复杂，本文就此作一归纳： 一、作副词，意为“相同地” ，“同样地”。例如： They don'thave as many airplanes. 他们没有同样多的飞机。 二、作连词， 1. 引导时间状语从句 as与when, while都是引导时间状语从句的从属连词，含义都是"当..... 的时候" 。但它们有区别: (1) . when作“当……的时候”解，可以指较短的(一点)时间,也可指一段时间。从句的动作可以与主句的动作同时发生，也可以先于主句的动作发生。例 如: John was having his dinner when I saw him. 当我看到约翰的时候，他正在吃饭。 She can write only when the baby is asleep. 只有婴儿睡着的时候，她才能写作。 (2) . while 常表示一段较长的时间或一个过程，强调主句谓语动词与从句谓语动词同时发生或在从句动作过程中发生。例如: We must strike while the iron is hot. 要趁热打铁。 While we were reading, the teacher came in. 我们正在读书的时候，老师走了进来。 (3) . 但属下列情形时，只用as, 而不用when 或while。 ①用于表示同一个人的两种动作交替进行，指“一边……，一边……”。例如: The girl dan ces as she si ngs on the stage那个女孩在舞台边唱歌边跳舞。 He looked behind from time to time as he went forward. 当他朝前走时，不时地向后看。 ②表示两个同步发展的动作或行为,译为“随着…… ”。例如: As time went on / by, she became more and more worried. 随着时间的流逝，她变得越来越焦虑。 As he grew older, he became more intelligent. 随着他年龄的增长，他变得更 有才智了。 ③表示两个短促行为或事件几乎同时发生。例如: I thought of it just as you opened your mouth. 恰好在你开口时，我想到了它。 Just as the flying worm hit her face, she gave a loud cry. 恰巧在飞虫撞到她脸上时，她大哭起来。 2. 引导原因状语从句 as，because, si nee都可以表示因果关系，连接原因状语从句，含义是“因为，由于”，但它们有区别：because表示的语气最强；as 一般放在句首，语气较弱，较口语化；since 常常用在书面语中，表示多为对方已知的、或稍加分析便可得知的原因，有时可译作“既然” 。例如： I do it because I like it. 我做这件事是因为我喜欢。 Since many of the customers work during the day, Billy has to collect the money at night. 因为许多顾客白天上班,所以毕利只好晚上去收钱。 As she has been ill perhaps she'll need some help. 她由于生病可能需要些帮